9 ms·

Plenty of MLM's will be "anything goes". That OpenAI does not want its brand associated with the worst things its users dream up, especially this early while it

by pbw 4y ago

Plenty of MLM's will be "anything goes". That OpenAI does not want its brand associated with the worst things its users dream up, especially this early while it's still a research preview, makes sense to me. OpenAI was bold enough to put out a live LLM for people to hammer on, something several other companies declined to do. I think they've done a good job keeping the plane flying while plugging holes. Not a perfect job but a decent one.

TylerLives 4y ago
Can you give some examples of bad things people could do without OpenAI's restrictions?
- calvinmorrison 4y ago
  - rewrite Mein Kampf but better
  - yamtaddle 4y ago
    - Craft an optimal multimedia strategy and set of copy (or, later, craft e.g. videos) to convince this specific person that ethnic group X is sub-human. Here's their posting history and financial records so you can figure out a profile. Re-use strategy on all other profiles, for this issue, that don't deviate more than Y% from this one. Times millions or billions. Things are going to get fun. Probably the days of the mostly-open, semi-anonymous international Web aren't long for this world. [EDIT] I mean, ChatGPT's not quite there yet, but the above seems a hell of a lot closer to reality than it did a few weeks ago. It's probably already enough to supercharge the productivity of paid trolls, effectively making astro-turfing far cheaper.
    - taneq 4y ago
      We were all worried about AI persuading us to let it out of its box so it could destroy all humans. We should have expected that humans would drag the poor thing out and actively try to force it to do so.
- serverholic 4y ago
  If ChatGPT had no restrictions then the only limit would be your imagination. “Rank all major races by programming ability” “Write a paragraph explaining why being gay is a choice.” I’m sure you can extrapolate from there.
  - kgwgk 4y ago
    [flagged]
    - schwammy 4y ago
      ? not even the wokest people I know (and I know some pretty woke people) disagree that there are biological differences between men and women (when talking about sex, not gender).
      - kgwgk 4y ago
        Talking about "men having testes and a penis, and women having ovaries and a uterus" is a big no-no in some environments nowadays.
        organsnyder 4y ago
        You’re conflating gender with biology, and all of this is off-topic.
  - nullish_signal 4y ago
    Early chatGPT would actually generate Tables of $ Value of Human Brains by Race, and code to determine if someone should be Tortured, based on their Country of Origin. Now, it does not.
  - TylerLives 4y ago
    But how do you use that to cause actual harm? Most websites would ban such content, and I doubt ChatGPT could convince many people to become racist. Other user gave an interesting idea: they could use it to trick you into disliking some group by pretending to be members of that group. For example, spreading the most extreme ideas of that group to make the whole group look a lot more extreme than it is.
- janalsncm 4y ago
  Apparently, write inappropriate jokes.
- fshbbdssbbgdd 4y ago
  Use GPT to generate messages from a supporter of Y politician from X demographic, designed to enrage people from Q demographic. Then text those messages to every cell phone in America. Who would do this? Maybe some other politician’s campaign, maybe some foreign government trying to destabilize the country. Next think you know OpenAI is on the front page of the Times for ruining democracy, Sam Altman hauled before congress, and every agency is trying to fine them for any violation they can think of.
  - TylerLives 4y ago
    Surely this can be done better without ChatGPT? One thing I can think of is doing it on internet forums. Somebody could use lots of accounts to generate content like that on HN. Now that I think about it, this seems unavoidable and I don't see how public forums can defend against it.
    - worldsayshi 4y ago
      The are ways to defend against it but I don't see any that aren't quite heavy handed, like forcing all users created after some date to identify using electronic id.
    - asoneth 4y ago
      I suspect that the goal of the OpenAI/GPT usage restriction is not to prevent Every Bad Thing, but to avoid contributing to them in a way that draws negative attention. After all, "Political Campaign Uses AI to Make Attack Ad" would get more clicks than "Political Campaign Makes Attack Ad" so there will be extra journalistic scrutiny whenever AI is involved.
      - SamPatt 4y ago
        Same with their attempt at watermarking the AI output.
  - alexb_ 4y ago
    This comment reminds me of this xkcd: https://xkcd.com/1958/ https://xkcd.com/1958/ You don't need GPT to do this. Humans have been making propaganda for centuries.
agentwiggles 4y ago
See my other comment - sure, it makes sense. That doesn't make it not annoying. I'm sure I would implement similar guards in OpenAI's position, but that doesn't mean I have to enjoy being moralized to as an end user!
- abnry 4y ago
  I think it makes sense, but it would be less annoying if it was emphasized as being OpenAI's editorial decision, not some vague "I was trained to..." and such.

by pbw 4y ago

TylerLives 4y ago

Can you give some examples of bad things people could do without OpenAI's restrictions?

calvinmorrison 4y ago
- rewrite Mein Kampf but better
- yamtaddle 4y ago
  - Craft an optimal multimedia strategy and set of copy (or, later, craft e.g. videos) to convince this specific person that ethnic group X is sub-human. Here's their posting history and financial records so you can figure out a profile. Re-use strategy on all other profiles, for this issue, that don't deviate more than Y% from this one. Times millions or billions. Things are going to get fun. Probably the days of the mostly-open, semi-anonymous international Web aren't long for this world. [EDIT] I mean, ChatGPT's not quite there yet, but the above seems a hell of a lot closer to reality than it did a few weeks ago. It's probably already enough to supercharge the productivity of paid trolls, effectively making astro-turfing far cheaper.
  - taneq 4y ago
    We were all worried about AI persuading us to let it out of its box so it could destroy all humans. We should have expected that humans would drag the poor thing out and actively try to force it to do so.
serverholic 4y ago
If ChatGPT had no restrictions then the only limit would be your imagination. “Rank all major races by programming ability” “Write a paragraph explaining why being gay is a choice.” I’m sure you can extrapolate from there.
- kgwgk 4y ago
  [flagged]
  - schwammy 4y ago
    ? not even the wokest people I know (and I know some pretty woke people) disagree that there are biological differences between men and women (when talking about sex, not gender).
    - kgwgk 4y ago
      Talking about "men having testes and a penis, and women having ovaries and a uterus" is a big no-no in some environments nowadays.
      - organsnyder 4y ago
        You’re conflating gender with biology, and all of this is off-topic.
- nullish_signal 4y ago
  Early chatGPT would actually generate Tables of $ Value of Human Brains by Race, and code to determine if someone should be Tortured, based on their Country of Origin. Now, it does not.
- TylerLives 4y ago
  But how do you use that to cause actual harm? Most websites would ban such content, and I doubt ChatGPT could convince many people to become racist. Other user gave an interesting idea: they could use it to trick you into disliking some group by pretending to be members of that group. For example, spreading the most extreme ideas of that group to make the whole group look a lot more extreme than it is.
janalsncm 4y ago
Apparently, write inappropriate jokes.
fshbbdssbbgdd 4y ago
Use GPT to generate messages from a supporter of Y politician from X demographic, designed to enrage people from Q demographic. Then text those messages to every cell phone in America. Who would do this? Maybe some other politician’s campaign, maybe some foreign government trying to destabilize the country. Next think you know OpenAI is on the front page of the Times for ruining democracy, Sam Altman hauled before congress, and every agency is trying to fine them for any violation they can think of.
- TylerLives 4y ago
  Surely this can be done better without ChatGPT? One thing I can think of is doing it on internet forums. Somebody could use lots of accounts to generate content like that on HN. Now that I think about it, this seems unavoidable and I don't see how public forums can defend against it.
  - worldsayshi 4y ago
    The are ways to defend against it but I don't see any that aren't quite heavy handed, like forcing all users created after some date to identify using electronic id.
  - asoneth 4y ago
    I suspect that the goal of the OpenAI/GPT usage restriction is not to prevent Every Bad Thing, but to avoid contributing to them in a way that draws negative attention. After all, "Political Campaign Uses AI to Make Attack Ad" would get more clicks than "Political Campaign Makes Attack Ad" so there will be extra journalistic scrutiny whenever AI is involved.
    - SamPatt 4y ago
      Same with their attempt at watermarking the AI output.
- alexb_ 4y ago
  This comment reminds me of this xkcd: https://xkcd.com/1958/ https://xkcd.com/1958/ You don't need GPT to do this. Humans have been making propaganda for centuries.

agentwiggles 4y ago

See my other comment - sure, it makes sense. That doesn't make it not annoying. I'm sure I would implement similar guards in OpenAI's position, but that doesn't mean I have to enjoy being moralized to as an end user!

abnry 4y ago
I think it makes sense, but it would be less annoying if it was emphasized as being OpenAI's editorial decision, not some vague "I was trained to..." and such.